Being Clear About AI: Our approach to the new EU Tranparancy Rules

Artificial intelligence is now part of everyday creative production. It helps us draft, edit, illustrate, narrate and analyse, and used well it makes good work better. It also raises a fair question that every client and every audience is entitled to ask: when was AI involved, and how would I know?

From 2 August 2026, the EU AI Act answers that question with law. Article 50 introduces transparency obligations for anyone producing or publishing AI-generated content professionally, and the European Commission has backed it with a Code of Practice on marking and labelling AI-generated content, published in June 2026. At Big Top Multimedia we think this is a welcome development, and we want to be upfront about where we stand.

What the new rules say

In plain terms, the AI Act now requires:

  • Clear labelling of deepfakes: AI-generated or AI-manipulated images, audio or video that could pass as authentic.
  • Labelling of AI-generated text published to inform the public on matters of public interest, unless a human has reviewed it and a person or organisation takes editorial responsibility.
  • Telling people when they are interacting with an AI system, such as a chatbot.
  • Machine-readable marking of AI outputs, a duty that sits with the companies who build the AI tools themselves.

The rules apply from 2 August 2026, with a grace period until 2 December 2026 for AI systems that were already on the market. Fines for getting it wrong can reach 15 million euro or 3 per cent of worldwide turnover. Honestly though, for us this was never about fines.

Where we stand

Our position is simple: nobody who sees, hears or reads our work should ever be misled about how it was made. Here is what that means in practice.

We tell you where AI was used. Every project record notes which AI tools contributed and how. If you are a client, you can ask at any time and get a straight answer.

A human is always responsible. Every piece of written content we deliver goes through human editorial review, with a named reviewer and clear editorial responsibility. AI can draft and assist; it does not get the final say.

We label what the law says to label, and a bit more. Synthetic voices in our podcast productions carry a spoken disclosure. Fully AI-generated imagery is labelled as such, and AI-modified photography is identified as modified. Where the EU’s new labelling icons suit the format, we use them.

We protect provenance data. Modern AI tools embed machine-readable marks in their outputs so that content can be verified. We preserve those marks through our production workflow wherever technically possible, and we never strip or disable them.

We help our clients comply. Most of our clients are the publishers of record for the work we create, which means these obligations land on them too. We deliver assets labelled correctly, with the documentation to back them up, and we advise on the right approach for each channel.

AI supports the craft; it does not replace it. Strategy, judgement, design sense and knowing an audience remain human work. We use AI where it genuinely improves the outcome, and we say so.

What this means if you work with us

Very little changes day to day, and that is the point. You will see clear AI disclosures on relevant deliverables, a short compliance note in our project documentation, and contract wording that spells out who is responsible for what at publication. If you publish content yourself, we will make sure you know exactly what labelling, if any, your organisation needs to apply.

For the professional bodies, charities, tourism organisations and public sector teams we work with, trust is the whole game. Being open about AI is not a burden on that trust. It is how you keep it.

Talk to us

If you are unsure how the AI Act’s transparency rules apply to your organisation’s content, we are happy to walk you through it, from podcast disclosures to website chatbots. Get in touch and we will keep it practical.

For more information about our approach to the use of AI, visit our AI Manifesto page.

This article is general information, not legal advice. For questions about your organisation’s specific obligations, talk to your legal adviser.

Ready when you are

Let us make something worth talking about.

Tell us what you are planning. One team, every medium, no hand-offs.